Guidance

General practice workforce estimates by constituency: background and methodology

Published 17 August 2023

Applies to England

Background

This publication presents modelled estimates of full time equivalent (FTE) clinical staff workforce operating in general practice in England. The data is presented at the start and end point of each time period for each constituency, and includes the constituency change over this period. It also shows the primary care networks (PCNs) active in each constituency. This publication covers:

• doctors in general practice
• nurses
• direct patient care (DPC) staff

We counted general practice clinical staff at individual practices to measure how many of these staff work within each constituency. Each practice is linked to the Parliamentary constituency it operates within and clinical staff totals are then grouped by constituency. We have used this approach for all timepoints for doctors and nurses in general practice and at the start timepoints for DPC staff.

At end timepoints of our comparisons, DPC staff are often employed directly by PCNs which can cover a number of practices. DPC staff employed by a PCN may carry out their work across practices within that PCN. We estimate that roughly half (47%) of PCNs operate across more than one constituency.

In order to assign PCN workforce to constituencies we have used a modelling methodology. We have modelled how these staff may be shared among practices by apportioning staff to practices based on their patient list size. We are then able to estimate DPC staff at constituency level for the end timepoints of our comparisons. Estimates are not required for our start points as PCNs were not established at that point.

Data sources

This publication uses publicly available data from NHS Digital and the Office for National Statistics (ONS). The sources used are:

• NHS Digital General Practice Workforce

• NHS Digital Primary Care Workforce Quarterly Update (experimental statistics)

• NHS Digital Payments to General Practice

• National Statistics Postcodes Lookup (NSPL)

Methodology

We assigned workforce to constituencies using a modelling methodology to apportion staff employed by PCNs to individual practices. We then added these estimates to the numbers of staff employed directly by practices. This methodology followed 4 steps:

1. The primary care workforce quarterly update collated figure provides the total number of DPC staff in general practice. This includes some DPC staff employed directly by practices who can be apportioned to constituencies through general practice workforce data. To avoid double counting, all staff included in general practice workforce data were first subtracted from the collated figure to leave only DPC staff employed directly by PCNs.

2. To determine how to allocate these PCN level staff to individual practices, we assigned a weight to each practice belonging to a PCN. We determined the practice weight by its weighted patient list size, which was taken from the payments to practices dataset. For example, if exactly half of a PCN’s weighted patient total belongs to one practice, that practice is assigned a weight of 0.5. The total weight for practices within each PCN totals 1. This approach provides an estimate based on notional demand for clinical staff based on the size of patient lists rather than an equal distribution across practices within a PCN.

3. To allocate staff employed directly by a PCN to practices, this weight is multiplied by the number of staff employed per PCN. If a practice with a weight of 0.5 operates within a PCN directly employing 10 DPC staff, this practice is allocated 5 DPC staff. To calculate our endpoint clinical workforce figures for each practice these DPC staff were added to practice-level DPC staff, doctors and nurses. The constituency total groups and sums the workforce of all practices within a constituency.

4. Our start point calculations do not include staff employed by PCNs because PCNs did not exist then. The start point clinical workforce figure instead directly groups and sums practice level staff from general practice workforce data for each constituency without the intermediate stages listed in steps 2 and 3. The change between our start and end timepoints was then calculated by subtracting the start point figure from the endpoint figure for each constituency.

Caveats and data limitations

Around half (47%) of PCNs span multiple constituencies and many DPC staff are employed by PCNs rather than practices due to the additional roles reimbursement scheme (ARRS). We have assigned these staff to practices according to the proportion of weighted patients a practice holds within a PCN. The real distribution of DPC staff is not expected to be as uniform as this assumption.

Data does not include estimates for practices that did not provide fully valid staff records.

DPC staff levels were calculated using full-time equivalent data. This refers to the proportion of full-time contracted hours that the post holder is contracted to work.

The data in this publication shows estimates of the workforce employed by practices and PCNs located in that constituency. Because constituencies are not a health geography, constituents may be registered with practices outside of their constituency. For that reason, this data is not a measure of the workforce that is available for constituents.

This approach produces an estimate of clinical staff for each constituency based on notional patient need. True staff levels are not known as not all staff record their activity at specific practices.

A small proportion of nurses and general practitioners are employed directly by PCNs rather than practices. These roles are not included in this data.

What this data can be used for

This approach allows us to estimate staff at constituency level, where this would not previously have been available. This is because constituencies are not a health geography.

The data uses a weighted patient method to estimate where staff would spend their working time. This approach generates estimates which are based on notional patient need rather than an equal division across practices or staff recording activity in specific practices. The data uses weighted patients rather than registered patients to better account for differences in patients between practices, such as age, which affect the clinical staff resource needed.

The dataset includes a lookup that shows the PCNs active in each constituency, so users can easily identify PCNs or constituencies of interest.